Weigang Lang

Title: Weigang Lang: Innovations in Data Processing and Parallel Computing

Introduction

Weigang Lang is an inventor based in Duluth, Georgia, whose work focuses on improving data loading and calculation processes through dynamic parallelization. He is associated with Oracle International Corporation and is credited with 2 patents.

Latest Patents

Lang’s patents address ways to divide complex data operations into coordinated tasks that can run in parallel.

“Dynamic parallelization of data loading” describes a system that scans a data file in segments and uses the scan results to determine how each segment can be loaded across multiple exclusive-access memory blocks. Loading proceeds in parallel with the scanning of subsequent segments, helping coordinate the two operations while respecting the information gathered during scanning.

“Dynamic parallelization of a calculation process” presents a method for creating a parallel calculation plan for rule sets or lists of rule-based expressions. The system divides expressions into task units, groups independent tasks at the same calculation level, and schedules those groups for execution across processing nodes. It can also split tasks into range-based subtasks, while preserving logical dependencies and the order needed for the overall calculation.

Career Highlights

Lang’s patent work reflects an emphasis on adapting parallel execution to the structure of a workload. His inventions describe systems that identify opportunities for concurrent processing while accounting for data boundaries, task dependencies, and available processing resources.

Collaborations

Lang’s listed coworkers include Jianwu Xu and Larry E. Roddenberry. His work is associated with Oracle International Corporation.

Conclusion

Weigang Lang’s two patents describe approaches to making data loading and rule-based calculations more efficient by dynamically organizing work for parallel execution.